No label defined (Q4725476)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 3998954
Language Label Description Also known as
default for all languages
No label defined
    English
    No label defined
    scientific article; zbMATH DE number 3998954

      Statements

      1987
      0 references
      prior
      0 references
      axioms for rational behavior
      0 references
      choice set
      0 references
      utility function
      0 references
      separability
      0 references
      conditional choices
      0 references
      0 references
      This important article provides what may be the weakest possible set of axioms for rational behavior. The axioms concern a choice set - a set of actions to be chosen by the agent. Or, importantly, a set of actions to be chosen by a collection of agents. We immediately have random combinations of choices, lotteries of lotteries, etc. From relatively innocuous (obvious, natural elementary) axioms concerning choice sets, we can establish the existence of a classical utility function. What we can not establish is the separability of the utility and the probability function. The author gives us a sufficient condition, in the form of an extra axiom, for making this distinction. In a certain sense, the axioms presented here provide a kind of set of minimal conditions enjoining the Bayesian view. This is clearly the way in which the author sees his axioms. Yet there are deep issues that are finessed by the author's approach. What he establishes are minimal conditions that entail static Bayesianism - i.e., the representation of an individual's doxastic and evaluative commitments at a given time, if he (or they - this is one important contribution here!) satisfies the constraints of the axioms, must be compatible with a Bayesian representation. But many think that the important part of Bayesianism lies in its dynamics. That is, if you are in a certain decision-theoretic state at time \(t_ 1\), and then evidence E comes in, you should be able to represent your new decision theoretic state in terms of a conditional probability and a utility function. The author's axioms have nothing to say on this score, though they may tell us how conditional choices should work. This is not to say there is anything wrong with the argument of this paper; it is only to say that we need more than these axioms to warrant the important dynamic part of Bayesianism.
      0 references

      Identifiers